[Bug-openmcl] Debugging crashes of dppccl

[Gary Byers]

On Mon, 8 Mar 2004, Erik Pearson wrote:

> Okay, attached is a file with the test script, and two diff files.
>
> To run the test (after patching)
>
> 1. Modify the host, path, timeout value or whatever within the file
> test-bug.cl, as you see fit.
> 2. compile and load the file "test-bug.cl"
> 3. run (test2 n) where n is the number of iterations. I usually use 10000
> to see how far it gets.
>
> Erik.
>

I was finally able to run this; thanks.

I once got a "PROCESS-INTERRUPT on exhausted thread" error; unless we
know a lot more about a thread than we probably do, we can't really
be sure that it won't exit before it acknowledges the interrupt; unless
it does a lot more handshaking than I think it wants to, PROCESS-INTERRUPT
probably has to be seen as something that -tries- to interrupt another
process.

I seem to be able to run a lot more iterations before crashing; I was
about to send this message saying that I wasn't able to get a
segfault, but I just got one (on the 12,273rd iteration.  In my case, it
looked like:

Unhandled exception 11 at 0x020f1ee0, context->regs at #xf01356c8
Read operation to unmapped address 0xf056000c
 While executing: #<Function %STACK-AREA-USABLE-SIZE #x060ef4de>
? for help
[2244] OpenMCL kernel debugger: Run 2273, result=(1089 NIL NIL NIL)
b


(#xF0135BB0) #x020F1B2C : #<Function INIT-THREAD-FROM-TCR #x060eefde> + 56
(#xF0135BC0) #x020F1A94 : #<Function NEW-LISP-THREAD-FROM-TCR #x060eef5e> + 72
(#xF0135BD0) #x020F6B1C : #<Function MAKE-PROCESS #x060f401e> + 216
(#xF0135BE0) #x020FE150 : #<Function %PROCESS-RUN-FUNCTION #x060ff89e> + 812
(#xF0135BF0) #x00004D7C : (subprimitive _ret1valn)
(#xF0135C00) #x020FE254 : #<Function PROCESS-RUN-FUNCTION #x060ff9be> + 128
(#xF0135C10) #x0637B79C : #<Function TEST2 #x0637b9de> + 328
(#xF0135C20) #x00004D7C : (subprimitive _ret1valn)
(#xF0135C30) #x0210DA34 : #<Function CALL-CHECK-REGS #x061120c6> + 72
(#xF0135C40) #x00004D7C : (subprimitive _ret1valn)
(#xF0135C50) #x0210C4D8 : #<Function TOPLEVEL-EVAL #x06110536> + 140
(#xF0135C60) #x0210CBCC : #<Function READ-LOOP #x06110af6> + 824
(#xF0135C70) #x0210CBE8 : #<Function READ-LOOP #x06110af6> + 852
(#xF0135C80) #x0210CC14 : #<Function READ-LOOP #x06110af6> + 896
(#xF0135C90) #x0210CC40 : #<Function READ-LOOP #x06110af6> + 940

Does that look at all familiar, or have the most recent crashes you've
seen been more random ?